To fix the banks, globalize them
Apr 30, 2010
3 minute read

As politicians and pundits debate what regulations are needed to fix the financial industry, I want to draw attention to a neglected issue: an important factor in the stability of the financial sector is its degree of globalization. All the big American banks have some international dealings. But true globalization of banking would be an improvement for both financial and political reasons.

The first and more trivial reason that banks should be more globalized is simply the logic of diversification. When investments are diversified, investors get higher return for less risk—it is a free lunch. Failure to internationally diversify is one of the biggest mistakes that American investors make (I’ve heard, as a rule of thumb, 30 percent of your investments should be foreign). But the big banks make this mistake as well! For instance, 90 percent of Bank of America’s revenues come from the domestic market. And of course, smaller banks frequently have no international holdings.

International diversification reduces systemic risk just as national diversification reduces systemic risk. If all banks could borrow from and lend to people in only a single town, we would see bank failures every time a factory shut down or a large business failed. By spreading that risk over thousands of towns, banks become more stable. But the logic of geographic diversification does not stop at national boundaries. In fact, because there are greater differences between countries than between towns in the same country, international diversification may be even more beneficial than national diversification.

The second and deeper reason for more financial globalization is political. The doctrine of Too Big to Fail creates a moral hazard problem: banks reap the benefits of successful investments but don’t suffer the losses of failed investments, so their incentive—whether conscious or evolved—is to take a lot of risk. Politicians, as much as they claim to want to enact market discipline, face a time-consistency problem. They want banks to be sound, but they lack the political will to let unsound banks and their creditors suffer when the time comes to feel the pain.

Greater globalization reduces the time-consistency problem. American voters do not want to pay taxes to bail out banks that operate just as extensively in Europe, Asia, and Latin America as in the US. The fact that a bailout would convey these uncompensated externalities would strengthen the political will to let unsound banks fail. This is not to say that I hope lots of banks fail—rather, I want politicians to be able to make more credible claims that Too Big to Fail is over, so that banks take note and invest accordingly. In equilibrium, there would be fewer bank failures, not more of them.

A number of commentators have argued that to deal with this political dimension of our financial troubles, we must break up the big banks. The argument is that if the banks were smaller, they would have less sway over politicians and the regulatory agencies. I have a lot of respect for many of the people making this argument, but I believe they are mistaken. One need only look at the sway that car dealerships, which are decentralized, have over state legislatures to see that small does not mean apolitical.

How can we achieve a more globalized financial sector? In truth, I don’t know. I suspect that only the insiders could say for sure. Perhaps we should ask them. At a minimum, we should consider repealing regulations that make it difficult or unprofitable for banks to operate internationally. Instead, regulatory reform will probably consist of a lot of rhetoric about getting tough with the banks—further proof that politics is not about policy.